Transcription
Journal of Information Systems Education, Vol. 21(2)The Importance of Emphasizing Individual Learning in the“Collaborative Learning Era”Aharon YadinRachel Or-BachManagement Information Systems DepartmentThe Academic College of Emek YezreelMobile mail Yezreel Valley 19300, ISRAELaharony@yvc.ac.il, orbach@yvc.ac.ilABSTRACTIn this paper we describe an instructional tactic of individually assigned homework that promotes and strengthens individuallearning processes. We claim that current emphasis on the benefits of collaborative learning belittles the importance ofindividual learning processes and reduces the opportunities to require and assess individual learning within IS education. Inour study, which used specially designed individual assignments, we succeeded in dramatically reducing the failure rate in twocourses in two consecutive semesters. We present findings from additional research tools that support and explain the changewe found in the failure rate when the tactic of the individually assigned homework was used. We conclude with someimplications of the suggested tactic including dealing with academic dishonesty and lowering the dropout rate in IS education.Keywords: Individual assignments, Individual homework, Individual learning, Effective learning1. INTRODUCTIONModern learning theories from any cognitive-constructivistparadigm assume that learning involves iterative processes ofstructuring, refining and restructuring of mental models.These processes are combined with other learning relatedprocesses like sense-making, debugging, evaluation,reflection and more. All these processes are necessary formeaningful learning whether employed in a context ofcollaborative or individual learning.When one examines the current published researchrelated to learning and particularly to computer-mediatedlearning, the proportion of research about collaborativelearning is astonishing. Even though proponents ofcollaborative learning acknowledge the important role ofindividual learning (Dillenbourg, 2005; Stahl, Koschmannand Suthers, 2006), current research papers deal mainly withcollaboration with very little mention of the individual facet.In addition, the research dealing with collaborative learningis shifting from looking at groups as a contextual variable toanalyzing group dynamics and looking at learning as a groupprocess. There is no doubt that collaborative learning hasmany advantages. There is also no doubt that groupdynamics is an important facet of collaboration, but there isno need to belittle the crucial facet of individual learning.As the focus of research influences practice and furtherresearch, we argue that more emphasis should be given toresearch regarding individual learning both as a prerequisiteand as a complementary facet of collaborative learning. Weargue further that as assessment tools shape and directstudents' learning goals, it is necessary to incorporate moreindividual assessment tools in higher education in order tofoster the necessity of individual learning skills andindividual accountability. That is not to say thatcollaboration is to be abandoned; on the contrary, ourargument has the goal of leveraging the benefits ofcollaborative learning processes. There is an underlyingimplicit assumption when dealing with collaborative learningprocesses that students are already used to learning asindividuals. It is an implicit assumption that students havealready practiced the relevant skills associated with learning,such as explaining to themselves, analyzing, synthesizing,combining and comparing to previous knowledge, makinggeneralizations, reflecting and other relevant skills. It seemsthat compared with the efforts given to investigating how tosupport collaborative learning, individual learning is notsupported enough. Even though collaborative learning can beseen as being the two facets of individual and group learningworking together, this does not imply that the best way topromote collaborative learning is by exercising collaborativelearning directly. We believe that there is much more need tofoster and assess individual learning in order to obtainmeaningful collaborative learning.In this paper we describe an instructional tactic forpromoting and strengthening individual learning processes.The instructional tactic suggested in this paper is based on aunique design for individually assigned homework. Byindividually assigned homework we mean homework that isrequired to be done individually (versus collaboratively).It isrequired to be done by the student him or herself, and185
Journal of Information Systems Education, Vol. 21(2)designed in such a way that each student uses different datathan the other students for performing the task. The ideabehind the design is to force students to try to employindividual learning processes. Intermediate and final valuesare different from one student to another and any comparison(or “borrowing”) of values is fruitless for completing thehomework assignments.The assignments are not dynamically adapted tostudents' characteristics and knowledge. There is no skillprofile or any use of student modeling capabilities as inintelligent tutoring systems. The individually assigned homework tactic that is described in this paper is much easier toimplement than more intelligent adaptation techniques, andstudents' intermediate inputs can be checked easily.The suggested tactic can also help in dealing withstudents' attitudes towards homework, in lowering studentdrop-out rates and in dealing with academic dishonestyamong students.2. LITERATURE REVIEWThe rationale for the design and implementation of theindividually assigned homework can be discussed in severalbroad contexts such as assessment or teaching strategies. Butin this literature review we focus on three contexts that relatemore specifically to our study and most importantly relate tocurrent trends. One is the relation between individuallearning and collaborative learning, another is self-efficacyand learning, and the last one is homework and academicdishonesty. Our aim is to show how these three contextsprovide the rationale for employing such a tool ofindividually assigned homework as suggested in this paper.2.1 Individual and Collaborative LearningResearch on learning in the last decades emphasizes theimportant role that collaborative learning plays in thelearning process. Collaboration is expected to promoteactivities like elaboration, justification and argumentationthat trigger learning mechanisms. Despite the expectations,there is no guarantee that these activities will occur withoutadditional educational design constraints (Dillenbourg, 1999).Information Technology graduates are expected to work inteams and collaboration skills are necessary; but how dotheir capabilities for individual work come in? Is it necessaryfor making the collaboration effective? Research on onlinecollaborative learning shows that for successful collaborativelearning to occur, students have to exhibit a high degree ofmotivation and involvement as well as both interdependenceand autonomy (Hansford and Wylie, 2002). In spite of themany benefits of the collaborative learning students still mayhave some problems using the method and display somedegree of unwillingness to participate in group learning(Barker, Garvin-Doxas, and Jackson, 2002; Waite et al.,2004). Morrison (2004) outlines another pitfall ofcollaborative learning and specifically collaborativeprogramming: the free riders. Free riders are students whoenjoy the benefits of collaborative work, but don’t contributeto the common goal. Joyce (1999) even defines the free-riderproblem as the biggest problem in collaborative learning.We believe that any successful collaboration starts withindividual capabilities and individual responsibility and186motivation. In this paper we stress the need for instructionaldesign for enhancing these individual capabilities, whichlater become a cornerstone in any collaboration activity.Some researchers dealing with instructional design forcollaborative learning also emphasize the individual facet(Puntambekar, 1999). Hoadley and Enyedy (1999) use themetaphor of monologue and dialogue to describe the socialactivities in which learning is grounded and suggest the needfor learning environments that help students’ transition fromdialogue to monologue and back again. Pair programming,for example, when employed as an instructionalmethodology emphasizes the different roles andresponsibilities of each participant. This collaborativeenvironment is effective only if each student carries his/herown task and does not “rely” on the other. This demonstratesthe importance of personal assignments and accountabilityeven in a collaborative framework. Within collaborativelearning research there are also studies where the conflictsbetween individual solutions are used to trigger effectivecollaborative learning (Constantino-Gonzalez, Suthers, andEscamilla, 2003; Or-Bach and Van Joolingen, 2004).We claim that there is not enough focus in the currentlearning research on ways to make students employ spirallearning processes by themselves: i.e. analyze, solve, debug,reflect, and repeat the process as long as necessary. Theseindividual capabilities (or learning habits) play a crucial rolein any future collaborative learning or collaborative workenvironments that the students will encounter.2.2 Self-efficacy and LearningDuring the past two decades, self-efficacy has emerged as ahighly effective predictor of students’ motivation andlearning. Self-efficacy is a person’s perception or judgmentof their own knowledge, capabilities, and capacity toperform a task at a specified level of performance (Bandura,1993; Seifert, 2004). Self-efficacy measures focus onperformance capabilities rather than on personal qualities,such as one’s physical or psychological characteristics.Respondents judge their capabilities to fulfill given taskdemands, such as solving fraction problems in arithmetic,not who they are personally or how they feel aboutthemselves in general. Self-efficacy beliefs are not a singledisposition but rather are multidimensional in form anddiffer on the basis of the domain of functioning(Zimmermann, 2000). Self-efficacy is essential for learning,since self-efficacy and motivation will influence efforts andvigor more than actual ability (Cavaco, Chettiar, and Bates,2003; Zusho, Pintrich, and Coppola, 2003). Students’judgment of their own self-efficacy in a discipline has beenfound to predict their performance in these disciplines(Glynn, Aultman, and Owens, 2005). Positive self-efficacyfor a task will lead to higher goals, more commitment, moreeffort and persistence. In addition, there is evidence thatstudents with positive self-efficacy beliefs are more likely tocontinue with even more difficult tasks (Linnenbrink andPintrich, 2002). Students with negative self-efficacy andbeliefs tend to give up when a task becomes difficult, or justavoid the task (Zimmermann, 2000). Research has verifiedthat self-efficacy is related positively to most of the factorsthat contribute positively to learning outcome: persistence,cognitive engagement, use of self-regulatory strategies and
Journal of Information Systems Education, Vol. 21(2)actual achievement (Bandura, 1997; Pintrich and Schunk,2002). Students should neither overestimate norunderestimate their capabilities; they should rather havefairly accurate, but optimistic beliefs about their efficacy toaccomplish a task (Linnenbrink and Pintrich, 2002).When it comes to pedagogical implications, selfefficacy is best facilitated by providing students with anopportunity to succeed. When students work withchallenging tasks within their range of competence,preferably towards short term goals, they strengthen theirpositive self-efficacy beliefs and at the same time developnew capabilities and skills (Glynn, Aultman, and Owens,2005). Instructors who give feedback should attempt tofoster positive but accurate self-efficacy beliefs. This is thechallenge for the design of homework, a design that relatesto content, submission procedures and assessment scheme.This challenge becomes more significant with current trends,as will be described in the following section.2.3 Homework and Academic DishonestyThere is a general agreement that homework plays animportant role in students’ learning. We argue that withoutexamining and re-examining the potential benefits ofhomework assignments and whether they are achieved, wemiss the opportunity to support students’ learning. This issuebecomes significantly important due to several trends inhigher education. Some of the trends relate to thecharacteristics of incoming students, and others to economicconstraints that affect the teaching load and the availabilityof teaching assistance. In many countries there has been atrend in the recent decade towards widening opportunitiesfor obtaining higher education. The result is that the studentpopulation gets more heterogeneous with regard to priorknowledge, learning habits, and cognitive and metacognitive skills that affect learning. The variety makes itnecessary for the teachers to have tools for formativeassessment and also makes it necessary for the students toexercise self-assessment. In a paper titled “Homework?What Homework?” (Young, 2002) the author summarizesfindings from the National Survey of Student Engagement ofthat year and suggests some explanations. “Students arestudying about one-third as much as faculty say they oughtto, to do well,” said the director of the survey. The moststriking statistic: Nineteen percent of full-time freshmen saythey spend only 1 to 5 hours per week preparing for classes.Many education experts say that is well below the minimumneeded to succeed. Seniors who answered the same surveyreported studying even less than freshmen, with 20 percentstudying 1 to 5 hours per week. Many professors say theirstudents are doing less homework these days, though thereare always a few model students. The problem may start inhigh school, where students are apparently spending far lesstime on homework than those who graduated a decade ago,and also have problems managing their time and getting themost out of their studying (Young, 2002).As many students come to higher education to makegood grades rather than explore new topics, academicdishonesty becomes prevalent. Academic dishonesty may bedefined as students’ attempt to present others’ academicwork as their own (Jensen et al., 2002). Academic dishonestyamong high school and college students is highly common—so common, in fact, that some observers describe it as an‘‘epidemic’’ (Haines et al., 1986). Academic dishonesty isnot a new problem, but it seems to get worse (Ercegovac andRichardson, 2004). Already in 1979, a Carnegie CouncilReport warned of ‘‘ethical deterioration’’ in academic life,and the U.S. Department of Education issued a reportdescribing cheating among college students as a ‘‘chronicproblem’’ (Maramark and Maline, 1993). The IEEETransactions on Education devoted a special issue in May2008 to the problem of plagiarism. The special issueincluded ten papers focusing on the topic of plagiarism. Themotivation behind the special issue was to uncover the rootcauses of plagiarism and suggest new ways of counteractingthese causes.When students submit homework assignments done byothers they miss the chance to learn, and the teacher missesthe chance to get a realistic mapping regarding students’understanding. As stated by Gibbs and Simpson (2004),plagiarism on assignments presents a serious problem for theintegrity of the educational process. Various tools weredeveloped for detecting plagiarism (Jones, 2008) andespecially for detecting plagiarism in programming courses(Zhang, Zhuang, and Yuan, 2007; Gitchell and Tran, 1999;Joy and Luck, 1999). Bowyer and Hall (2001) in their paperabout reducing effects of plagiarism in programming classesdescribe the effectiveness of such a system – MOSS(Measure Of Software Similarity). They further stress thatdetection of program plagiarism is made relatively simpleusing MOSS but the real challenge for the faculty member isto design procedures that reduce the perceived pressure onstudents to cheat and make the learning process moreeffective. Our approach is along similar lines; we are notinterested in punishing students and even though we try toraise ethical issues, still our main goal is to maintain aneffective educational process. The approach we suggest inthis paper is not an approach for detecting plagiarism afterthe fact, but an approach for designing assignments thatmake plagiarism more difficult and thus support students’learning. Study results of Broeckelman-Post (2008) showedthat students’ engagement in academic dishonesty is mostinfluenced by whether they believe their peers are engagingin academic dishonesty. This is a good reason to invest in thedesign of assignments that explicitly require individual workand make plagiarism more difficult.3. INDIVIDUALLY ASSIGNED HOMEWORK ANDTHE RESPECTIVE COURSESThe research described in this paper was conducted withintwo courses: (1) Computer Organization and Programming;and (2) Systems Architecture. This section provides a briefdescription of these two courses, an example of an individualassignment and a further explanation of the instructionaltactic of individually assigned homework. Another example,along with a description of the initial use of the individuallyassigned homework in the Computer Organization andProgramming course, can be found in a previous paper(Yadin and Or-Bach, 2008).187
Journal of Information Systems Education, Vol. 21(2)3.1 The Computer Organization and ProgrammingCourseThe Computer Organization and Programming (COAP)course is a mandatory, introductory course which is intendedto provide basic understanding of computer systemoperations, data representation, system architecture andAssembly language. The participating students are in theirsecond year. The course is aimed at software developers andits main objective is to enhance the students’ understandingof hardware functions and operations. The Assemblylanguage is used to enable students to demonstrate theirunderstanding of the various hardware components. Thiscourse is a pre-requisite for the Systems Architecture course.3.2 The Systems Architecture CourseThe Systems Architecture (SA) course is an elective secondyear course mainly for students who are looking to improvetheir knowledge regarding the technology used in the variousinformation systems solutions. This course is intended toenhance students’ knowledge regarding basic hardwarefunctionality, and new technological developments andpossible impacts they may have on organizationalinformation systems solutions.In both courses a student’s final grade is calculatedbased on a final exam (70%), a mid-term exam (20%) andseveral (at least 6) homework assignments (10%).We had these courses running for several years with ongoing evaluation and respective modifications. The studentsconsidered these courses to be difficult ones and the failurerate was disturbing. The courses were accompanied by anaction research study that brought up some changes in thecourses over the years, such as the inclusion of mid-termexams, additional in-class lab exercises and revisedassignments, both manual and computerized. Despite theimprovement attempts there was a constant increase in thefailure rate percentage, consistent with the decrease inenrollment. During the academic year 2007-2008 weintroduced into these courses the idea of individualassignments. We tried to foster individual learning bydesigning assignments that make students invest more timein the task before comparing with other students as they areused to doing.All the assignments in the above described courses wereof the “individualized” type. Each submitted assignment wasgraded and in addition, since feedback is essential for thestudents’ improvement, detailed informative feedback wasprovided. The feedback included extra explanations (whenneeded), and links to the learning materials and to additionalexercises. Our electronic submission system was used topublish the assignments and set the last date for submission,to collect the students’ work and to present to each studentthe relevant feedback for each submitted assignment.3.3 An Example of an Individual AssignmentThe following is an example of an individual assignmentgiven in the COAP course. The purpose of the assignment isto assess understanding of the [Segment:Offset] concept andthe hardware stack mechanism.a.Absolute Addressing1. On top of the assignment write your 9 digit studentID number (N1N2N3N4N5N6N7N8N9)1882. Starting from the left-hand side, divide the IDnumber into groups of 3 digits each (N1N2N3N4N5N6 N7N8N9)3. Calculate the Binary equivalence of the number ineach of the groups.4. Assume that the rightmost group is the Segmentaddress and each of the other groups representsdifferent offsets.5. Calculate the absolute addresses referred to by theseoffsets. (N7N8N9:N1N2N3N7N8N9:N4N5N6)b.Stack Addressing and Content1. Write once again your student ID number2. Starting from the left-hand side divide the ID numberinto groups of 2 digits each (0N1 N2N3 N4N5 N6N7N8N9)3. Calculate the Binary equivalence of the number ineach of the groups.4. Assume that the rightmost group (N8N9) representsthe Stack Segment starting address and the StackOffset.5. Each of the other groups represents values to beentered into the Stack.6. Write down the absolute addresses and the Stackcontent after executing the following assemblerinstructions:PUSH 0N1PUSH N2N3PUSH N4N5PUSH N6N7This type of assignment requires the students tocarefully analyze the algorithm principles and then tomentally execute it. The mental execution helps studentsunderstand the abstract algorithm and provides the student aswell as the teacher with evidence regarding thisunderstanding. The use of individual input data for executingthe algorithm ensures that each student follows all the stepsof the algorithm. This type of assignment makes itimpossible to “import” the full or partial solution from acolleague or compare results before employing selfmonitoring/debugging procedures. Any help provided by afellow student or a teaching assistant has to concentrate onthe solving process without mentioning exact outcomes. Thisis again a measure to make students practice by themselvesthe cognitive processes required for meaningful learning.The individual assignments provide a good mechanismfor assessing the students’ knowledge by closely analyzingtheir intermediate answers. For this specific assignment,evaluating students’ understanding at an early stage of thecourse was very important since the hardware stack in thex86 architecture works in a peculiar way (as the top of stackpointer decreases the stack actually grows). Based onfeedback accumulated in previous years, the stack proved tobe a difficult point for students to grasp. While working onthe assignment, the students had to demonstrate theirunderstanding by applying the stack principles to theirindividual data. The assignment relates to both the stackcontent as well as addressing behavior including dealingwith end cases such as stack overflow/underflow. In theevent of an erroneous reply, the student got back his/herassignment including feedback that directly related to the
Journal of Information Systems Education, Vol. 21(2)specific error. Sometimes an explanation was added,including the directing of the student to the relevant sectionin the learning materials.The “individualization” method just described mightalso have an affective effect, making students more attachedand motivated to solve their own tasks. In this case studentsmight relate better to any feedback given to them becausethey feel that the feedback is personal – relevant to their“own” problem and produced especially for them. Since thestudents think about their assignment by themselves, thefeedback they receive makes sense to them.4. THE STUDY – TOOLS AND RESULTS4.1 IntroductionIn order to investigate the effect of the individually assignedhomework we used several research tools. The main tool wasthe comparison of students’ failure percentage during theyears that these courses were taught. We also administered apost-course survey to the students who used the individuallyassigned homework in order to better understand the resultswe got from the failure percentage data. Two other researchtools were also used to explain and cross validate the resultsof the failure percentage comparison. These tools werecomparison of students’ access to the Learning ManagementSystem during the study year and the year before, andinformal interviews with some students.The individual assignments were introduced for the firsttime in the academic year 2007-2008. In the ComputerOrganization and Programming (COAP) course during theacademic year 2007-2008 there were 18 students (39%female and 61% male) and in 2008-2009 there were 27students (22% female and 78% male).In the Systems Architecture course during the academicyear 2007-2008, there were 14 students (57% female and43% male) and in 2008-2009 there were also 14 students(29% female and 71% male).4.2 Failure PercentageCompleting the course successfully requires passing theexam and then the final score is calculated by the specificscheme for the final score of that course. As was mentionedin the courses’ description, in both courses a student’s finalgrade was calculated based on a final exam (70%), a midterm exam (20%) and several (at least 6) homeworkassignments (10%).The following figures describe the failure percentage ofboth courses during the years that these courses wereadministered. The years in the graphs are an abbreviation ofthe academic year, where for example 2009 means theacademic year 2008-2009. Figure 1 describes the failurerates for the COAP course, while figure 2 describes thefailure rates for the SA course. In both figures the number ofstudents who took the course during this year appears inparentheses under the year.Both figures show a clear change of trend since the newtactic of individual assignments was introduced in 20072008. The academic year 2007-2008 was the first year everthat no one failed the Systems Architecture course, as can beseen in figure 2. This was repeated in 2008-2009 as well.During the 2005-2006 academic year, the SA course was notoffered, so in the graph we used the average of 2004-2005and 2006-2007. In the Computer Organization andProgramming course the decrease in failure percentage isalso dramatic, as can be seen in figure 1: In 2006-2007(before the introduction of the new method) it was 14.3%;later in 2007-2008 it dropped to 5.6%; and in 2008-2009 itdropped to 3.7%.The numbers of students indicated in thetwo graphs show the decrease in the number of studentsduring these years. This decrease could have providedanother explanation for the reduction in the failure rate. But amore careful examination shows that in the COAP course in2004-2005 there were 23 students and the failure rate was8.7%, while in 2008-2009 there were 27 students with afailure rate of 3.7%. Similarly, in 2005-2006 there were (16)2007(42)Figure 1: Failure rate in the COAP course.1892008(18)2009(27)
Journal of Information Systems Education, Vol. 2006(0)2007(17)2008(14)0.0%2009(14)Figure 2: Failure rate in the SA course.students with a failure rate of 12.5%, while in 2007-2008there were 18 students with a failure rate of 5.6%. So thenew method of the individually assigned homework seems amore plausible explanation for the decrease in failure rate.4.3 The SurveyA survey was designed in order to examine the students’attitudes towards the individual assignments. The survey wasadministered in a subsequent semester in order to get fromthe students a retrospective view after they had finished thecourse and taken the final test. A translation of the surveyfrom Hebrew appears in appendix 1.The survey had 17 Likert type items. The items relatedto facts (“ I devoted more time ”), opinions (“Helping afellow student in the individual assignments method is morechallenging than helping with other learning methods”),feelings (“Due to the use of individual assignments I feltmore prepared for the final exam”), beliefs (“Ibelieve increase motivation ”), preferences (“I preferindividual assignments instead of the kind of assignmentsused in other courses”) and wishes (“I’d like to haveindividual assignments in all courses”). The scale was 1-5,where 1 was “totally disagree” and 5 was “completelyagree”.Students were also asked to summarize in free text theiropinion about the individual assignments and to add anycomments or suggestions they had regarding the individualassignment method.Fourteen students filled in this survey. These were thestudents who studied both courses during the academic year2008-2009, so their input represents their attitude based ontwo courses in two semesters where the individualassignment method was employed.4.4 Survey AnalysisWe calculated the average score for each of the survey items.If we exclude item 11, which does not relate directly to theemployment of the individual assignments, we see thatstudents are in favor of this method. The average of the190averages (excluding item 11) is 3.71. The highest averagewas for item 4: “With the individual assignments I felt theinstructor comments addressed my own work”. The averagescore for this item was 4.79 with standard deviation of 0.58.This is a very interesting finding as it means that studentsexpect and appreciate the attention to their individual work.In the free text this was also clearly expressed by one of thestudents: “Getting feedback adapted to me led me to i
many benefits of the collaborative learning students still may have some problems using the method and display some degree of unwillingness to participate in group learning (Barker, Garvin-Doxas, and Jackson, 2002; Waite et al., 2004). Morrison (2004) outlines another pitfall of collaborative learning and specifically collaborative
